Method and device for determining a frequency mask for a frequency spectrum

A method for automatic determination of a frequency mask disposed above or below a frequency spectrum of a detected signal by a computer device, the method comprising selecting a tolerance value, selecting an ordinate value disposed at the left or right end of the frequency spectrum as a first ordinate value of the frequency mask and selecting by successively determining a respectively next ordinate value of the frequency mask as an ordinate value of the frequency spectrum which is disposed at a furthest distance from a respectively last-determined ordinate value of the frequency mask, wherein ordinate values of the frequency spectrum disposed between the current and the last-determined ordinate value of the frequency mask each provides a vertical spacing distance relative to a connecting line between the current and the last-determined ordinate value of the frequency mask that is smaller than the selected tolerance value, and

• determining an optimized approximation function for the characteristic of the ordinate values of the frequency mask in the environment of local maxima and/or local minima by:
• identifying local maxima or minima in the individual ordinate values of the frequency mask,
• determining all ordinate values of the frequency mask in an environment of the respective local maximum or of the respective local minimum by determining the ordinate values of the frequency mask between the respective maximum and a threshold value positioned below the respective maximum at a given spacing distance or between the respective minimum and a threshold value positioned above the respective minimum at a given spacing distance, and
• determining a function type and function parameters of the approximation function by minimizing a metric between the ordinate values of the frequency mask and the ordinate values of the approximation function in the environment of the respective maximum or the respective minimum.

A device for determining a frequency mask disposed above or below a frequency spectrum of a detected signal, the device comprising:

• a computer;
• a first functional unit including instructions for execution by the computer to determine ordinate values of a first envelope curve disposed completely above or below the frequency spectrum,
• a second functional unit including instructions for execution by the computer to determine ordinate values of a second envelope curve disposed completely above or below the frequency spectrum and completely above or below the first envelope curve, and
• a third functional unit including instructions for execution by the computer to determine ordinate values of a frequency mask disposed between the first and second envelope curve, wherein in each case two successive ordinate values of the frequency mask with a maximum horizontal spacing distance within the first and second envelope curve can be reached in a linear manner relative to one another, and
• an approximator for approximating a noise floor through ordinate values of the frequency spectrum disposed on a horizontal straight line.

A device for determining a frequency mask disposed above or below a frequency spectrum of a detected signal, the device comprising:

• a computer;
• an eighteenth functional unit including instructions for execution by the computer to select an ordinate value disposed at the left or right edge of the frequency spectrum as the first ordinate value of the frequency mask,
• a nineteenth functional unit including instructions for execution by the computer to successively determine the respectively next ordinate value of the frequency mask, and
• a twentieth functional unit including instructions for execution by the computer to generate the frequency mask disposed above or below the frequency spectrum of the detected signal by weighting the determined ordinate values of the frequency mask, wherein the characteristic points of the approximation function are selected from the group consisting of maximum, maximum with the deduction of a first threshold value, maximum with the deduction of a threshold value or respectively minimum, minimum with the deduction of a first threshold value, minimum with the deduction of a threshold value in the case of a parabolic or Gaussian approximation function, and four corner points of a trapeze-shaped approximation function.